Real Industrial Environment Research Articles

Low-Cost UWB Based Real-Time Locating System: Development, Lab Test, Industrial Implementation and Economic Assessment.

This paper presents the technical development and subsequent testing of a Real-Time Locating System based on Ultra-Wideband signals, with the aim to appraise its potential implementation in a real industrial case. The system relies on a commercial Radio Indoor Positioning System, called Qorvo MDEK1001, which makes use of UWB RF technology to determine the position of RF-tags placed on an item of interest, which in turn is located in an area covered by specific fixed antennas (anchors). Testing sessions were carried out both in an Italian laboratory and in a real industrial environment, to determine the best configurations according to some selected performance indicators. The results support the adoption of the proposed solution in industrial environments to track assets and work in progress. Moreover, most importantly, the solution developed is cheap in nature: indeed, normally tracking solutions involve a huge investment, quite often not affordable above all by small-, medium- and micro-sized enterprises. The proposed low-cost solution instead, as demonstrated by the economic assessment completing the work, justifies the feasibility of the investment. Hence, results of this paper ultimately constitute a guidance for those practitioners who intend to adopt a similar system in their business.

JujubeNet: A high-precision lightweight jujube surface defect classification network with an attention mechanism.

Surface Defect Detection (SDD) is a significant research content in Industry 4.0 field. In the real complex industrial environment, SDD is often faced with many challenges, such as small difference between defect imaging and background, low contrast, large variation of defect scale and diverse types, and large amount of noise in defect images. Jujubes are naturally growing plants, and the appearance of the same type of surface defect can vary greatly, so it is more difficult than industrial products produced according to the prescribed process. In this paper, a ConvNeXt-based high-precision lightweight classification network JujubeNet is presented to address the practical needs of Jujube Surface Defect (JSD) classification. In the proposed method, a Multi-branching module using Depthwise separable Convolution (MDC) is designed to extract more feature information through multi-branching and substantially reduces the number of parameters in the model by using depthwise separable convolutions. What's more, in our proposed method, the Convolutional Block Attention Module (CBAM) is introduced to make the model concentrate on different classes of JSD features. The proposed JujubeNet is compared with other mainstream networks in the actual production environment. The experimental results show that the proposed JujubeNet can achieve 99.1% classification accuracy, which is significantly better than the current mainstream classification models. The FLOPS and parameters are only 30.7% and 30.6% of ConvNeXt-Tiny respectively, indicating that the model can quickly and effectively classify JSD and is of great practical value.

Predicting the Impact of Product Type Changes on Overall Equipment Effectiveness Through Machine Learning

Nowadays, Industry 4.0 and the Smart Manufacturing environment are increasingly taking advantage of Artificial Intelligence. There are more and more sensors, cameras, vision systems and barcodes in the production area, as a result of which the volume of data recorded during manufacturing and assembly operations is growing extremely fast. The interpretation and processing of such production-type data by humans is no longer possible effectively. In the Big Data domain, machine learning is playing an increasingly important role within data mining. This paper focuses on the product change processes of semi-automatic assembly line batch production and examines the impact of product type changes on the Overall Equipment Effectiveness (OEE) and attempts to determine future values through supervised machine learning. Using decision tree technology, the effect on the OEE value can be predicted with an accuracy of up to 1%. The presented data and conclusions come from a real industrial environment, so the obtained results are proven in practice.

Industrial Fault Detection using Contrastive Representation Learning on Time-series Data

Deep learning (DL) has been known as one of the effective techniques for building data-driven fault detection methods. The successful DL-based methods require the condition that massive labeled data are available, but this is sometimes an inevitable obstacle in real industrial environments. As one of the solutions, autoencoders (AEs) are widely adopted since AEs can extract features from unlabeled data. However, some challenges in AE-based fault detection methods remain, such as the design of encoder architecture, the computational cost, and the usage of the limited labeled data. This paper proposes a new industrial fault detection method through learning instance-level representation of time-series based on the self-supervised contrastive learning framework (SSCL). The proposed method uses dilated-causal-convolution-based encoder-only architecture to extract the information from industrial time-series data. A new data augmentation method for time-series data is proposed based on the temporal distance distribution, which is used to construct positive pairs in SSCL. Moreover, the encoder is alternately trained by the new weighted contrastive loss and the traditional classification loss. Finally, the experiments are conducted on the industrial data set and a semi-physical system, showing the effectiveness of the proposed method.

Welding defects classification through a Convolutional Neural Network

This letter presents a Convolutional Neural Network (CNN), named WelDeNet, customized to classify welding defects, such as lack of penetration (LP), cracks (CR), porosity (PO) and no defect (ND), by inspecting digitalized radiographic images. A new dataset that collects 24,407 images representing welding defects is also presented. WelDeNet consists of 14 cascaded convolutional layers and achieves a test accuracy of 99.5 %. When hardware implemented within the Raspberry Pi 3B + board, WelDeNet exhibits an inference time of only 134 ms, with CPU and memory utilizations of just 51 % and 47 MB, thus offering a promising solution easy-to-integrate in a real industrial environment.

An adaptive compliance Hierarchical Quadratic Programming controller for ergonomic human–robot collaboration

This paper proposes a novel Augmented Hierarchical Quadratic Programming (AHQP) framework for multi-tasking control in Human–Robot Collaboration (HRC) which integrates human-related parameters to optimize ergonomics. The aim is to combine parameters that are typical of both industrial applications (e.g. cycle times, productivity) and human comfort (e.g. ergonomics, preference), to identify an optimal trade-off. The augmentation aspect avoids the dependency from a fixed end-effector reference trajectory , which becomes part of the optimization variables and can be used to define a feasible workspace region in which physical interaction can occur. We then demonstrate that the integration of the proposed AHQP in HRC permits the addition of human ergonomics and preference. To achieve this, we develop a human ergonomics function based on the mapping of an ergonomics score, compatible with AHQP formulation. This allows to identify at control level the optimal Cartesian pose that satisfies the active objectives and constraints, that are now linked to human ergonomics. In addition, we build an adaptive compliance framework that integrates both aspects of human preferences and intentions, which are finally tested in several collaborative experiments using the redundant MOCA robot. Overall, we achieve improved human ergonomics and health conditions, aiming at the potential reduction of work-related musculoskeletal disorders. • Human–robot interaction and collaboration in manufacturing industrial applications. • Hierarchical control improves robot flexibility in real industrial environments. • Hierarchical human ergonomics improves posture, lowering musculoskeletal diseases. • An adaptive compliance framework introduces human preferences alongside ergonomics. • Optimal trade-off between productivity- and human-related tasks.

AN INNOVATIVE TECHNOLOGY FOR PRODUCTION OF AXIAL INDUCTION MOTOR

This study presents the latest technology to produce axial induction motors for application in electric vehicles. A review of current state-of-the-art technology is performed on the solutions with soft magnetic powder core or machined parts that are not appropriate for mass production. Therefore, new technology is proposed, based on analyzed production technologies specifics. It involves use of a standard metal strips to make both stator and rotor cores. This design includes a spirally wound strip of magnetically conductive material and stacked trapezoidal lamellas. The trapezoidal lamellae are cut using waste-free technology for e.g., guillotine, without specialized punches as in traditional engines. As a result, metal cutting operations are significantly simplified and require widely used production equipment in the industry. The developed technology is appropriate for mass production due to its relatively low cost and eases adaptability. The proposed design is at the physical prototyping and testing stage. Thus, it will be validated by testing the physical prototype in the real industrial environment. This innovative technology is patented and currently under commercialization.

Asymmetric HMMs for Online Ball-Bearing Health Assessments

The degradation of critical components inside large industrial assets, such as ball-bearings, has a negative impact on production facilities, reducing the availability of assets due to an unexpectedly high failure rate. Machine learning-based monitoring systems can estimate the remaining useful life (RUL) of ball bearings, reducing the downtime by early failure detection. However, traditional approaches for predictive systems require run-to-failure (RTF) data as training data, which in real scenarios can be scarce and expensive to obtain as the expected useful life could be measured in years. Therefore, to overcome the need of RTF, we propose a new methodology based on online novelty detection and asymmetrical hidden Markov models (As-HMMs) to work out the health assessment. This new methodology does not require previous RTF data and can adapt to natural degradation of mechanical components over time in data-stream and online environments. As the system is designed to work online within the electrical cabinet of machines, it has to be deployed using embedded electronics. Therefore, a performance analysis of As-HMM is presented to detect the strengths and critical points of the algorithm. To validate our approach, we use real life ball-bearing data sets and compare our methodology with other methodologies where no RTF data are needed and check the advantages in RUL prediction and health monitoring. As a result, we showcase a complete end-to-end solution from the sensor to actionable insights regarding RUL estimation toward maintenance application in real industrial environments.

Condensation Heat Transfer Correlation for Micro/Nanostructure Properties of Surfaces.

Condensation, which can be observed in nature as a phase change heat transfer phenomenon, is a critical phenomenon in industrial fields such as power generation, water desalination, and environmental control. Many existing studies have applied surfaces with different wettability by controlling the surface topology to enhance condensation heat transfer. However, the industrial applicability is close to zero due to the limited size and shape of surfaces and low supersaturation conditions. Here, we regulate the surface topology of large-area copper tubes, which are representative industrial metals. We fabricated four copper tubes with different surface structures. We analyzed the condensation phenomenon of the modified tube under specific supersaturation conditions by measuring the overall heat transfer coefficient. We analyzed the condensation phenomenon by measuring the condensation heat transfer coefficient. We have recognized that there is a difference between the maximum droplet radius and the droplet detaching frequency depending on the size and shape of the structure. We measured the contact angle and contact angle hysteresis to accurately analyze the droplet behavior on each surface. As a result, we show that there is a correlation between contact angle hysteresis (CAH) and the total heat transfer coefficient, indicating heat transfer performance. These findings can be applied when evaluating surfaces with excellent condensation heat transfer performance for use in real industrial environments, which can dramatically reduce time and cost.

Few shot cross equipment fault diagnosis method based on parameter optimization and feature mertic

With the rapid development of industrial informatization and deep learning technology, modern data-driven fault diagnosis (MIFD) methods based on deep learning have been receiving attention from the industry. However, most of these methods require sufficient training samples to achieve the desired diagnostic effect, and the scarcity of fault samples in the actual industrial environment leads to the limitation of the development of MIFD methods. In addition, data-driven fault diagnosis methods often need to face cross-load or even cross-domain problems across different devices due to changes in equipment operating conditions and production requirements. In this paper, we design a parameter optimization and feature metric-based fault diagnosis method with few samples, called model unknown matching network model, for the problem of sparse fault samples and cross-domain between data sets in real industrial environments. The method combines both a parametric optimization-based meta-learning network, which extracts optimization information to adapt between different domains, and a metric-based metric learning network, which extracts metric information for similarity discriminations. The experimental results show that the method outperforms the current baseline method for the five-shot fault diagnosis problem of bearings under limited data conditions and achieves an accuracy of up to 94.4 in cross-device diagnosis experiments from bearings to gas regulators, indicating the feasibility of the method. The features are visualized by T-SNE to show the validity of the model.

SMoCo: A Powerful and Efficient Method Based on Self-Supervised Learning for Fault Diagnosis of Aero-Engine Bearing under Limited Data

Vibration signals collected in real industrial environments are usually limited and unlabeled. In this case, fault diagnosis methods based on deep learning tend to perform poorly. Previous work mainly used the unlabeled data of the same diagnostic object to improve the diagnostic accuracy, but it did not make full use of the easily available unlabeled signals from different sources. In this study, a signal momentum contrast for unsupervised representation learning (SMoCo) based on the contrastive learning algorithm—momentum contrast for unsupervised visual representation Learning (MoCo)—is proposed. It can learn how to automatically extract fault features from unlabeled data collected from different diagnostic objects and then transfer this ability to target diagnostic tasks. On the structure, SMoCo increases the stability by adding batch normalization to the multilayer perceptron (MLP) layer of MoCo and increases the flexibility by adding a predictor to the query network. Using the data augmentation method, SMoCo performs feature extraction on vibration signals from both time and frequency domains, which is called signal multimodal learning (SML). It has been proved by experiments that after pre-training with artificially injected fault bearing data, SMoCo can learn a powerful and robust feature extractor, which can greatly improve the accuracy no matter the target diagnostic data with different working conditions, different failure modes, or even different types of equipment from the pre-training dataset. When faced with the target diagnosis task, SMoCo can achieve accuracy far better than other representative methods in only a very short time, and its excellent robustness regarding the amount of data in both the unlabeled pre-training dataset and the target diagnosis dataset as well as the strong noise demonstrates its great potential and superiority in fault diagnosis.

Anomaly detection as vision-based obstacle detection for vehicle automation in industrial environment

Nowadays, produced cars are equipped with mechatronical actuators as well as with a wide range of sensors in order to realize driver assistance functions. These components could enable cars’ automation at low speeds on company premises, although autonomous driving in public traffic is still facing technical and legal challenges. For automating vehicles in an industrial environment a reliable obstacle detection system is required. State-of-the-art solution for protective devices in Automated Guided Vehicles is the distance measuring laser scanner. Since laser scanners are not basic equipment of today’s cars in contrast to monocameras mounted behind the windscreen, we develop a computer vision algorithm that is able to detect obstacles in camera images reliably. Therefore, we make use of our well-known operational design domain by teaching an anomaly detection how the vehicle path should look like. The result is an anomaly detection algorithm that consists of a pre-trained feature extractor and a shallow classifier, modelling the probability of occurrence. We record a data set of a real industrial environment and show a robust classifier after training the algorithm with images of only one run. The performance as an obstacle detection is on par with a semantic segmentation, but requires a fraction of the training data and no labeling.

HT-TPP: A Hybrid Twin Architecture for Thermal Power Plant Collaborative Condition Monitoring

Thermal power plants, TPP, are one of the main players in the phosphoric acid and fertilizer production value chain. The control of power plant assets involves considerable complexity and is subject to several constraints, affecting the asset’s reliability and, most importantly, plant operators’ safety. The main focus of this paper is to investigate the potential of an agent-based digital twin architecture for collaborative prognostic of power plants. Based on the ISO 13374:2015 scheme for smart condition monitoring, the proposed architecture consists of a collaborative prognostics system governed by several smart DT agents connected to both physical and virtual environments. In order to apprehend the potential of the developed agent-based architecture, experiments on the architecture are conducted in a real industrial environment. We show throughout the paper that our proposed architecture is robust and reproduces TPP static and dynamic behavior and can contribute to the smart monitoring of the plant in case of critical conditions.

The finite element modeling of the impacting process of hard particles on pump components

Abstract Shield pumps which transport a variety of complex media are widely used in chemical industry, mineral industry, and other fields. However, the hard particles in the media erode the flow passage in the shield pump, seriously affecting the service life of the pump. Therefore, it is important to study the erosion process in the flow passage to extend the service life of the pump. This erosion process in the shield pump can be investigated by experiments, but it is difficult for the current experimental method to copy the real industrial service environment of impellers. For example, the environment temperature and speed range are found difficult to be reached by the current experimental method. Comparatively, the simulation by finite element method can overcome the above deficiency. In this article, the ANASYS software is used to simulate the erosion process in shield pump flow passage due to hard particles. Specifically, the structural static analysis module is used to build the model of thermal barrier coating components and hard particle components. Then the variation law of stress, strain, and material deformation of the flow passage components under concentrated force is obtained. Second, the extended finite element method is used to study the crack propagation caused by the erosion process. A linear elastic–linear softening constitutive model is established to simulate the stress variation during the crack propagation and material removal under the continuous erosion from hard particles. The proposed law regarding stress variation and crack propagation in the erosion process in this study contributes to theoretical support for damage detection and service life extension of shield pumps.

Hydrogen Isotope Separation Using a Metal–Organic Cage Built from Macrocycles

AbstractPorous materials that contain ultrafine pore apertures can separate hydrogen isotopes via kinetic quantum sieving (KQS). However, it is challenging to design materials with suitably narrow pores for KQS that also show good adsorption capacities and operate at practical temperatures. Here, we investigate a metal–organic cage (MOC) assembled from organic macrocycles and ZnII ions that exhibits narrow windows (<3.0 Å). Two polymorphs, referred to as 2α and 2β, were observed. Both polymorphs exhibit D2/H2 selectivity in the temperature range 30–100 K. At higher temperature (77 K), the D2 adsorption capacity of 2β increases to about 2.7 times that of 2α, along with a reasonable D2/H2 selectivity. Gas sorption analysis and thermal desorption spectroscopy suggest a gate‐opening effect of the MOCs pore aperture. This promotes KQS at temperatures above liquid nitrogen temperature, indicating that MOCs hold promise for hydrogen isotope separation in real industrial environments.

Hydrogen Isotope Separation Using a Metal-Organic Cage Built from Macrocycles.

Porous materials that contain ultrafine pore apertures can separate hydrogen isotopes via kinetic quantum sieving (KQS). However, it is challenging to design materials with suitably narrow pores for KQS that also show good adsorption capacities and operate at practical temperatures. Here, we investigate a metal–organic cage (MOC) assembled from organic macrocycles and ZnII ions that exhibits narrow windows (<3.0 Å). Two polymorphs, referred to as 2α and 2β, were observed. Both polymorphs exhibit D2/H2 selectivity in the temperature range 30–100 K. At higher temperature (77 K), the D2 adsorption capacity of 2β increases to about 2.7 times that of 2α, along with a reasonable D2/H2 selectivity. Gas sorption analysis and thermal desorption spectroscopy suggest a gate‐opening effect of the MOCs pore aperture. This promotes KQS at temperatures above liquid nitrogen temperature, indicating that MOCs hold promise for hydrogen isotope separation in real industrial environments.

Design and Implementation of Smart Glasses with ISMS and Risk Management Functionalities for a Technical Operator

This work aims to describe the design and implementation of smart glasses to be used by industrial operators allowing the display environment related information such as temperature, humidity as well as machine maintenance operations, the associated risks, and Information Security Management System (ISMS) controls. In order to achieve this design, several analyses were conducted in order to determine the optimal design possible namely an external and an internal functional and a risk analysis, in addition to an ergonomic study. Based on the aforementioned study and analyses, a model was created and simulated using the commercial software SoildWorks and ISISPORTEUS. The smart glasses functionalities have been evaluated in a real industrial environment.

Intrusion Detection System After Data Augmentation Schemes Based on the VAE and CVAE

Industrial Internet of Things (IoT) is the most rapidly developing industry in the current IoT industry, and the intrusion detection system (IDS) remains one of the key technologies for industrial IoT security protection. Researchers have considered applying algorithms such as machine learning and deep learning to network IDSs to cope with complex and changing network environments and to automatically extract key features from high-dimensional feature data. However, in the real industrial IoT environment, data imbalance is the main factor that affects the performance of the deep-learning-based IDS. In this article, we study the network intrusion detection model based on data level. Three data-based research schemes are constructed step by step in this article, which are a data augmentation scheme based on the variational autoencoder (VAE), a data-balancing scheme based on the conditional VAE, and a data-balancing scheme based on random undersampling and conditional VAE. The three data-level-based schemes are combined with the deep-learning-based IDS. In this article, we build experiments based on the CSE-CIC-IDS2018 dataset to verify the effectiveness of three data processing schemes. After data enhancement through the third scheme, the Macro-F1-score of the convolutional-neural-network-based IDS model improved by 3.75% and the Macro-F1-score of the gated-recurrent-unit-based IDS model improved by 5.32%.

Long-range Dependencies Learning Based on Non-Local 1D-Convolutional Neural Network for Rolling Bearing Fault Diagnosis

In the field of data-driven bearing fault diagnosis, convolutional neural network (CNN) has been widely researched and applied due to its superior feature extraction and classification ability. However, the convolutional operation could only process a local neighborhood at a time and thus lack ability of capturing long-range dependencies. Therefore, building an efficient learning method for long-range dependencies is crucial to comprehend and express signal features considering that the vibration signals obtained in a real industrial environment always have strong instability, periodicity, and temporal correlation. This paper introduces non-local mean to the CNN and presents 1D non-local block (1D-NLB) to extract long-range dependencies. The 1D-NLB computes the response at a position as a weighted average value of the features at all positions. Based on it, we propose a non-local 1D convolutional neural network (NL-1DCNN) aiming at rolling bearing fault diagnosis. Furthermore, the 1D-NLB could be simply plugged into most existing deep learning architecture to improve their fault diagnosis ability. Under multiple noise conditions, the 1D-NLB improves the performance of the CNN on the wheelset bearing dataset of high-speed train and the Case Western Reserve University bearing dataset. The experiment results show that the NL-1DCNN exhibits superior results compared with six state-of-the-art fault diagnosis methods.

Multi-Stream Convolutional Neural Networks for Rotating Machinery Fault Diagnosis under Noise and Trend Items.

In recent years, rotating machinery fault diagnosis methods based on convolutional neural network have achieved much success. However, in real industrial environments, interfering signals are unavoidable, which may reduce the accuracy of fault diagnosis seriously. Most of the current fault diagnosis methods are of single input type, which may lead to the information contained in the vibration signal not being fully utilized. In this study, theoretical analysis and comprehensive comparative experiments are completed to investigate the time domain input, frequency domain input, and two types of time–frequency domain input. Based on this, a new fault diagnosis model, named multi-stream convolutional neural network, is developed. The model takes the time domain, frequency domain, and time–frequency domain images as input, and it automatically fuses the information contained in different inputs. The proposed model is tested based on three public datasets. The experimental results suggested that the model achieved pretty high accuracy under noise and trend items without the help of signal separation algorithms. In addition, the positive implications of multiple inputs and information fusion are analyzed through the visualization of learned features.